<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
wav2vec2_xlsr50k_english_timit_phoneme
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the TIMIT dataset. It achieves the following results on the evaluation set:
- Loss: 0.5842
- Cer: 0.1560
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 4.8072 | 6.94 | 500 | 0.9883 | 0.4153 |
| 0.566 | 13.88 | 1000 | 0.3517 | 0.1535 |
| 0.3564 | 20.83 | 1500 | 0.3760 | 0.1570 |
| 0.272 | 27.77 | 2000 | 0.4290 | 0.1562 |
| 0.2131 | 34.72 | 2500 | 0.4919 | 0.1574 |
| 0.1667 | 41.66 | 3000 | 0.5499 | 0.1564 |
| 0.1359 | 48.61 | 3500 | 0.5842 | 0.1560 |
Framework versions
- Transformers 4.11.3
- Pytorch 1.12.1
- Datasets 1.13.3
- Tokenizers 0.10.3
